This is a systematic presentation of the divine, human, animal and natural (plants, rivers, clouds, etc.) figures in early Sanskrit literature (mainly the Rig Veda, the Atharva Veda, and the early Brahmanas), with the small number of narratives which actually appear in the early hymns, and some information from medieval Sanskrit commentators. (Technically, some of this is in pre-Sanskrit Vedic, not the systematized Sanskrit of the Indian grammatical tradition....) The much more elaborate narratives of the Epics and Puranas are not treated as such, and it is NOT a presentation of Hinduism as a living religion; those who want either one will be disappointed. The result is in many cases a list of epithets and attributes, and of family members, enemies, and rivals, at considerable remove in style from the lyrics from which the information has been gathered. It is a place to start looking for data, and is a guide to some (hardly all) of the contents of an extremely difficult body of ancient literature. Although written in English, it was originally published in 1897-1898 as a part of a German publisher's series of monographs on what would now be called Indo-European Studies. Macdonnell was a leading scholar in his time, and his technical studies continue to be cited with respect. The book is now showing its age, and a replacement would be welcome, but as its second century begins it remains extremely helpful. (At least to someone who wants a concise overview of a huge amount of scattered material.) The proportion of hard data to theory is very high. Macdonnell's work was able to survive changing fashions, and, to judge from citations and bibliographies, was used happily by rival schools of interpretation right through the twentieth century. A rather heavy dependence on nature-mythology (especially solar) seems to show direct dependence on the great commentator Sayana as least as much as on Max Mueller's modern extensions of it, as one would expect of the author's mastery of the sources. The Sanskrit index is quite comprehensive, and supplements the analytical arrangement of the material. Unfortunately, it lacks an index of texts cited. (In frustration at flipping pages to see if Macdonnell had mentioned a passage which did not have a major name, nor a word for which I somehow happened to know the Sanskrit, I once compiled my own index of his Rig Veda citations. I do not recommend this course to others!) The bibliographical material is, of course, long obsolete. The cautious reader will try to check Macdonnell's etymologies against modern works. He seems to have been fairly careful about accepting speculations not grounded in Sanskrit texts (not rushing to identify Vata and Wodan, for example), but of course, even the great Sanskrit grammarians were sometimes wrong....